K-Means Clustering based Lexicon Analytical Model for Multi-Source News Classification

The supervised models have been found more efficient for the purpose of news classification. The major goal of the news classification research is to improve the accuracy while decreasing the elapsed time. It is always difficult for the people to read all of the news on their favourite’s portal which have listed over the given portal. In this research, an approach is KNN lexicon technique which is used to obtain the popular news list from thousands or hundreds of online news available through APIs. This approach uses extraction summarization for summarizing the keywords thereby selecting the original sentences and putting it together into a new shorter text explaining the overall overview of the news data. Then the lexicon analysis would be performed over the given text data and then final classification of the news is done using k-nearest neighbor. The results would be obtained in the form of the parameters of accuracy, elapsed time, etc.